Essays and practice notes for legal professionals — legal writing, research technique, and the honest, practical side of lawyering in a fast-changing world.
Comprehensive research database of 1,624+ AI hallucination sanctions cases worldwide, $110,000 Oregon penalty (U.S. record), $86,000 Florida sanctions, technical framework, ABA Opinion 512 compliance, and evidence-based verification protocols. ~45 min read.
Read research →Comprehensive research analysis of the AI-powered pro se litigation surge — complete case database, empirical litigation data, defense cost analysis, technical architecture, and the unauthorized practice of law framework. ~40 min read.
Read research →No blanket duty exists — but five situations require disclosure, state bars disagree on the rest, and courts are imposing their own requirements.
Read →Research-grade analysis with complete sanctions case law database, Stanford empirical studies (17–88% hallucination rates), technical framework, and evidence-based verification protocols. ~22,000 words, 35 min read.
Read research →The desk-level guide — what the opinion actually requires for competence, confidentiality, candor, supervision, and billing when you use AI, with the documentation that proves it.
Read →The doctrine, the first ruling on point, and what makes the expectation of confidentiality reasonable when AI sees client material.
Read →A fair comparison of AI legal research tools against Westlaw and Lexis — what AI genuinely adds, where the incumbent databases still win, the measured accuracy gap, and how a hybrid workflow looks.
Read →A fast, reliable workflow for checking whether an AI-generated case citation is real — existence, holding, treatment, context — with free tools and the red flags that catch fabrications.
Read →The retrieval pipeline behind legal AI research tools — keyword and semantic search, reranking, grounding — and why the retrieval step decides whether the answer can be trusted.
Read →The specific ways AI contract review fails in practice — misclassification, missed cross-references, misgrounded flags — and the checklist of what still needs a human eye.
Read →Turn your firm's judgment into reusable, AI-checkable rules — the anatomy, governance, and a starter playbook you can adapt.
Read →The clauses lawyers fight over most — the risk flags that matter, the COVID-era cases, and where AI flagging stops and judgment begins.
Read →OCR, chunking, hybrid search, clause classification, and playbook rules — the pipeline between upload and verdict, explained for lawyers.
Read →Clause extraction, risk flags, and compliance playbook scanning — plus the honest limits where human judgment is irreplaceable.
Read →ABA Formal Opinion 512, competence and supervision duties, and a five-step verification workflow that survives scrutiny.
Read →Adoption statistics show AI replaces tasks, not judgment. What the redistribution of legal work means for your career.
Read →The three ways client data can leak, what "private AI" really means, and how local processing protects confidentiality.
Read →Why legal AI hallucinates fake citations, how grounded retrieval fixes it, and the verification habit that separates a tool from a malpractice risk.
Read →Why this blog exists, what you'll find here, and how these notes are written — plainly, for lawyers who value their time and their craft.
Read →